The Food and Drug Administration (FDA) has not provided effective oversight for drugs approved on the basis of indirect measurements of efficacy, according to a recent report by the Government Accountability Office (GAO).

Intermediary endpoints are sometimes used in place of explicit clinical measurements of efficacy in the FDA's review of new drug applications. For example, HIV/AIDS drugs are often approved based on their ability to reduce viral load rather than on reductions of clinical complications. Anticancer treatments may be approved without data on survival rate if they demonstrate tumor-growth inhibition.

The use of indirect endpoints allows drug companies to conduct smaller, shorter, and thus less expensive clinical trials to win marketing approval. It also allows the FDA to review and approve new, possibly superior treatments for serious or life-threatening diseases through an expedited process that was implemented in 1992.

Indirect endpoints are also acceptable for the nonexpedited review process if these measurements have well-established correlations with clinical benefits. For example, a new drug can be approved for treating type 2 diabetes if it is shown to reduce blood glucose levels, even in the absence of data on reduced clinical complications.

The risk of approving new drugs based on indirect measurements is that the inferred effectiveness does not always translate into prolonged lives or reduced clinical events, the GAO report pointed out. One such case was the anticancer drug Iressa, which was approved in 2003 for treatment of non-small-cell lung cancer based on tumor shrinkage in a minority of patients. The drug then failed to prolong survival in a postmarketing study. The FDA left the drug on the market but added severe restrictions on its use.

In this report, released in September, the GAO reviewed drugs approved based on indirect endpoints from 1992 to December 2008 and concluded that the FDA has not been diligently tracking postmarketing studies the agency itself had requested.

Although drug companies are required to file annual updates on each of their postmarketing studies until the FDA deems the case closed, the agency does not routinely and promptly review these annual reports, because this monitoring is considered of low priority, the GAO report said. As a result, the agency cannot keep track of the status or progress of ongoing postmarketing studies. It is difficult for the agency, or any interested party, to find out whether a company is meeting its study obligations. The GAO discovered that the FDA very rarely sends letters to companies that fail to file annual updates on time.

Even when the FDA does find that companies have failed to meet postmarketing requirements, it often takes no action. The agency has never withdrawn a drug from the market because a company failed to comply with the postmarketing requirement or to confirm the drug's effectiveness in postmarketing studies, FDA officials told the GAO. GAO auditors cited several cases in which drugs met the criteria for withdrawal but remained on the market.

The FDA approved 90 new drug applications from 1992 to 2008 based on indirect endpoints in expedited reviews, the GAO found. Over one-third of the 144 postmarketing studies required to confirm drug efficacy remained incomplete.

The report also identified 69 drugs approved by the FDA based on well-established endpoints through the nonexpedited review process. Although these drugs are not required to confirm efficacy in postmarketing studies, 175 postmarketing studies have been requested by the FDA for other reasons. The report noted that only half of these studies were completed.